Preventing car battery fires with help from machine learning
University of Arizona Research Enhances Electric Vehicle Battery Safety
New research led by University of Arizona doctoral student Basab Goswami is paving the way for improved safety in electric vehicles by predicting and preventing dangerous battery overheating. Published in the Journal of Power Sources, Goswami’s paper introduces a groundbreaking framework that combines multiphysics and machine learning to detect and prevent thermal runaway in lithium-ion batteries.
Supported by a $599,808 grant from the Department of Defense, Goswami and his adviser, Professor Vitaliy Yurkiv, have developed an advanced system that uses thermal sensors and historical data to forecast potential overheating events. This innovative approach aims to be integrated into electric vehicle battery management systems, potentially averting catastrophic failures and enhancing driver safety.
The research comes at a pivotal time as the electric vehicle market surges, with a 35% increase in global sales in 2023 and significant investments from the Biden administration. Goswami’s work addresses critical safety concerns, offering a lighter and more cost-effective solution compared to previous methods involving thermal imaging.